Skip to main content

Deploying with Docker Compose

While it is easy to evaluate Flexify.IO functionality by creating a free account at manage.flexify.io, or deploy Flexify.IO from cloud marketplace, in many cases you may need an on-premises installation.

Components​

Flexify.IO installation consists of:

  • Management Server - holds configuration data and provides REST APIs for management and monitoring.

  • SQL database (MariaDB by default) - stores configuration and monitoring data.

  • Web Console - provides browser-based GUI for management and monitoring.

  • One or more Engines that process S3 application requests and data migrations.

All components can be consumed from the cloud as Flexify.IO SaaS offering, installed on-premises, in your cloud account, or mixed in a hybrid solution (for example, on-premises engines managed by cloud-based management server).

Single-Machine Installation​

For evaluation and PoC purposes it is recommended to configure all the components on a single machine.

  1. Provision a physical or virtual/cloud machine:

    • Minimum specs: 1 CPU, 2 GB RAM

    • Recommended specs: 2 CPU, 4 GB RAM

    • Recommended OS: Ubuntu 26.04 LTS

  2. Download and run the installation script:

    curl -Of 'https://download.flexify.io/docker/latest/install.sh'
    sudo bash install.sh -v latest

    Note: the installation script will install Docker and Docker Compose, create services configuration and an empty configuration database, and start Docker containers.

  3. Open the following URL in a web browser to access Flexify.IO management UI: <http://<ip-address>> or <http://<ip-address>>/flexify-io/manage/

  4. Create an account using any email address as a username.

    Note: Flexify.IO does not have any “call home” features and will not send any personal information anywhere outside of your organization.

After creating an account, you will be brought to the management console where you can add on-premises and cloud object storages, configure virtual endpoints and virtual buckets for multi-cloud storage and run data migrations.

Production Installation​

For better performance, scalability, and reliability in production environments we recommend:

  • Run at least two engines on dedicated machines in each region where data is accessed from, optionally behind a load balancer.

  • Back up the configuration database.

  • Replicate the database if your RTO and RPO requirements cannot be achieved by restoring the database (or the entire VM) from a backup (engines and data flow will remain fully operational in case Management Server is temporarily unavailable).

Install Management Server​

  1. Provision a physical or virtual/cloud machine:

    • Minimum specs: 1 CPU, 2 GB RAM

    • Recommended specs: 2 CPU, 4 GB RAM, or higher if you plan to manage over 50 engines.

    • Recommended OS: Ubuntu 26.04 LTS

  2. Download and run the installation script:

    curl -Of 'https://download.flexify.io/docker/latest/install.sh'
    sudo bash install.sh -v latest --multi-engine
  3. Open the following URL in a web browser to access Flexify.IO management UI: <http://<ip-address>> or <http://<ip-address>>/flexify-io/manage/

  4. Create an account using any email address as a username.

    Note: Flexify.IO does not have any “call home” features and will not send any personal information anywhere outside of your organization.

Install Engines​

  1. Provision two or more Ubuntu 26.04 LTS machines with at least 1 GB RAM.

    Note: we recommend scaling horizontally rather than vertically, i.e., adding new machines with engines rather than adding CPU or RAM to existing machines.

  2. Log in to Flexify.IO Management at <http://<ip-address>> and open the All Engines tab.

  3. Click Add Engine.

  4. Copy-paste the command from the Ubuntu / CentOS section to a terminal on each of the provisioned machines, replacing your_company and your_region with actual information that can be later used for engines identification.

    Note: it is also possible to provision engines as Docker containers.

  5. Engines should now appear in the list on the All Engines tab and are ready to process data.

  6. Optionally configure a load balancer of your choice to distribute inbound S3 traffic among engines:

    • Inbound: TCP ports 80 and 443 (the load balancer needs to handle SSL)

    • Outbound: any of the engines at port 80

    • Affinity: none, engines are stateless, i.e. any engine can process any request

Change Configuration Database​

Flexify.IO uses MariaDB (default) or Microsoft SQL Server (configurable) to store configuration information (we do not store actual objects or objects catalogue). By default, the installation script will run MariaDB in a Docker container using Docker volume flexify-sql to store data.

For production environments you may want to consider:

  • Backing up the configuration database. Refer to MariaDB or Microsoft documentation on how to back up and restore MariaDB or Microsoft SQL Server database in a Docker container.

  • Using a custom Microsoft SQL Server installation (potentially shared with other users) managed by the company’s IT department or a dedicated DBA team.

To switch to Microsoft SQL Server:

  1. Create an empty database and enable snapshot isolation with the following SQL commands:

    ALTER DATABASE <database-name> SET ALLOW_SNAPSHOT_ISOLATION ON
    ALTER DATABASE <database-name> SET READ_COMMITTED_SNAPSHOT ON
  2. Edit /etc/flexify/docker-compose.yml:

    • Set address and credentials of your Microsoft SQL Server database:

      - SPRING_DATASOURCE_URL=jdbc:sqlserver://<ip:port>;database=<database-name>;
      - SPRING_DATASOURCE_USERNAME=<SQL username>
      - SPRING_DATASOURCE_PASSWORD=<SQL password>
    • Remove the sql service.

    • Remove the dependency on sql from the management-server service.

  3. Apply changes:

    sudo -E docker-compose up -d --remove-orphans

The Management Server will create the necessary database structure at the first start.

Next Steps​

Once installed, you may use Flexify.IO to: